Scribd
Upload
88 views4 pages

Fourier Transform

The document discusses the Fourier transform as it relates to continuous-time aperiodic signals with finite duration or energy. It introduces the Fourier transform pair, where the inverse transform synthesizes the signal from its spectrum and the direct transform provides the analysis of the signal's spectrum. The Fourier transform represents the spectrum as continuous rather than discrete lines, with integration rather than summation. The Fourier transform exists for signals that are absolutely integrable or have finite energy, which covers nearly all signals encountered in practice.

Uploaded by

Istiaque Ahmed
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
88 views4 pages

Fourier Transform

The document discusses the Fourier transform as it relates to continuous-time aperiodic signals with finite duration or energy. It introduces the Fourier transform pair, where the inverse transform synthesizes the signal from its spectrum and the direct transform provides the analysis of the signal's spectrum. The Fourier transform represents the spectrum as continuous rather than discrete lines, with integration rather than summation. The Fourier transform exists for signals that are absolutely integrable or have finite energy, which covers nearly all signals encountered in practice.

Uploaded by

Istiaque Ahmed
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

240 Frequency Analysis of Signals and Systems Chap.

it becomes an energy signal and its average power is zero. The spectra of finite energy
signals are described in the next section.
We also note that if k # 0 and sin(rrkF,,s) = 0. then cr = 0. The harmonia
with zero power occur at frequencies kF0 such that r r ( k F o ) r = mrr, m = f 1, f2, . . .,
or at k Fo = m / r . For example, if Fo = 4 Hz and T = 0.27,. it follows that the spectra]
components at f20 Ifi, 1 4 0 H z , . . . have zero power. These frequencies correspond
to the Fourier coefficients CL. k = f 5 . 110, 115. . . . . On the other hand. if r = O.lTp,
the spectral components with zero power are k = f10, f20, k30.. ...
The power density spectrum far the rectangular pulse train is

4.1.3 The Fourier Transform for Continuous-Time


Aperiodic Signals

In Section 4.1.1 we developed the Fourier series to represent a periodic signal


as a linear combination of harmonically related complex exponentials. As a con-
sequence of the periodicity, we saw that these signals possess 11ne spectra with
equidistant lines. The line spacing is equal to the fundamental frequency, which
in turn is the inverse of the fundamental period of the signal. We can view the
fundamental period as providing the number of lines per unit of frequency (line
density). as illustrated in Fig. 4.6.
With this interpretation in mind, it is apparent that if we allow the period to
increase without limit, the line spacing tends toward zero. In the limit, when the
period becomes infinite, the signal becomes aperiodic and its spectrum becomes
continuous. This argument suggests that the spectrum of an aperiodic signal will
be the envelope of the line spectrum in the corresponding periodic signal obtained
by repeating the aperiodic signal with some period T,.
Let us consider an aperiodic signal x ( r ) with finite duration as shown in
Fig. 4.7a. From this aperiodic signal. we can create a periodic signal x , ( t ) with pe-
riod T,, as shown in Fig. 4.7b. Clearly. x , ( t ) = x ( t ) in the limit as T, -+ oo,that is,

This interpretation implies that we should be able to obtain the spectrum of x(t)
from the spectrum of x,(r) simply by taking the limit as T, 4 m.
We begin with the Fourier series representation of x , ( r ) ,

where
Sec. 4.1 Frequency Analysis of Continuous-Time Signals

-7 , -T , , 0 T,,/2 742 I
Figure 4.7 ( a ) A p e r ~ o d i csignal .I ( I )
and ( h ) periodic signal x p ( t ) constructed
ih) by repeatinp x ( t ) with a period 7,,.

Since x - , ( I ) = x ( t ) for -Tp/2 5 r 5 Tp/2. (4.1.21) can be expressed as

It is also true that .t ( r = 0 Cor It > T,,/2. Consequentl!~. the limits on the intesral
in (4.1.22) can bc replaced by - x and x.Hence

Let us now define a function X ( F ) , called the Fourier transform of x ( t ) ,as

X ( F )= 1%
x
r (t)e-~'"~'dr (4.1.24)

XCFl is a function of the continuous variable F. It does not depend o n T, o r


Fo. However, if we compare (4.1.73) and (4.1.24). it is clear that the Fourier
coefficients c~ can be expressed in t e r n s of X ( F ) as

o r equivalently.

Thus the Fourier coefficients are samples of X ( F ) taken at multiples of Fo and


scaled by Fo (multiplied by l/T,,). Substitution for ck from (4.1.25) into (4.1.20)
yields
242 Frequency Analysis of Signals and Systems Chap. 4

We wish to take the iimit of (4.1.26) as Tp approaches infinity. First. we define


A F = l / T p . With this substitution. (4.1.26) becomes

It is clear that in the limit as T,,p approaches infinity, x,(r) reduces to x ( r ) . Also, AF
becomes the differential d F and k AF becomes the continuous frequency variable
F. In turn. the summation in (4.1.27) becomes an integral over the frequency
variable F. Thus

lim x,,(I) = x ( I ) = lim


Tp-x *F+Ok=-x
r Ci

x ( ~ A F ) ~ J A~F " ~ ~ '

x(r) = 1- 3L'
x(F)~J~""~F

This integral relationship yields x ( t ) when X ( F ) is known, and it is called the


inv erse Fourier transform.
This concludes our heuristic derivation of the Fourier transform pair given
by (4.1.24) and (4.1.28) for an aperiodic signal x ( t ) . Although the derivation is
not mathematically rigorous, it led to the desired Fourier transform relationships
with relatively simple intuitive arguments. In summary, the frequency analysis of
continuous-time aperiodic signals involves the folIowing Fourier transform pair.

FREQUENCY ANALYSIS O f CONTINUOUS-TIME APERIODIC SIGNALS

Synthesis equation
inverse transform

Analysis equation
dtrect transform

It is apparent that the essential difference between the Fourier series and the
Fourier transform is that the spectrum in the latter case is continuous and hence
the synthesis of an aperiodic signal from its spectrum is accomplished by means of
integration instead of summation.
Finally. we wish to indicate that the Fourier transform pair in (4.1.29) and
(4.1.30) can be expressed in terms of the radian frequency variable 52 = 2 n F .
Since d F = d!2/27r . (4.1.29) and (4.1.30) become

~ ( r=) -1
I
2~ =,.-
=
X ( R ) ~ J ~ ~ S (4.1.31)
c€
X(R) =
1, ~(r)e-~"dr

The set of conditions that guarantee the existence of the Fourier transform is the
(4.1.32)
Sec. 4.1 Frequency Analysis of Continuous-Time Signats

Dirichler condirions, which may be expressed as:


1. T h e signal .r(r) has a finite number of finite discontinuities.
2. The signal x ( r ) has a finite number of maxima and minima.
3. The signal x ( r ) is absolutely integrable, that is.

(4.1.33)
X

The third condition follows easily from the definition of the Fourier transform,
given in (4.1.30). Indeed.

Hence 1 X ( F ) I < x if (4.1.33) is satisfied.


A weaker condition for the existence of the Fourier transform is that x ( r )
has finite energy: that is.
LXCC
i.~-(!)i~dt< x

N o ~ ethat if a signal x(i) is absolutely integable. it will also have finite energy.
That is.
r
,.r(r)ldr < x
Z
then

However. the converse is not true. That is. a signal may have finite energy but
may not be absolutely integrable. For example, the signal
sin 2rr r
x(r) = -
T!
is square integrable but is not absolutely integrable. This signal has the Fourier
transform

Since this signal violates (4.1.33). it is apparent that the Dirichlet conditions are
sufficien~but not necessary for the existence o f the Fourier transform. In any case.
nearly all finite energy signals have a Fourier transform. so that we need not worry
about the pathological signals, which are seldom encountered in practice.

4.1.4 Energy Density Spectrum of Aperiodic Signals

Let x(r) be any finite energy signal with Fourier transform X ( F ) . Its energy is

Ex = 1 3C

Y:
ix(r)12dr

You might also like